基于Matlab的发酵过程平台软件设计

Design of Optimal Software in Ferment Process Based on Matlab

设计了一套优化软件,将先进的智能控制理论运用到发酵过程控制领域中。针对非线性、时变的发酵过程利用神经网络非线性预测控制方法,建立了发酵过程神经网络模型进行菌体浓度、基质糖浓度、产物浓度的在线估计。由于缺乏发酵过程精确的数学模型,采用神经网络非线性预测控制方法结合遗传算法寻优技术确定发酵过程控制参数的最佳值,使整个发酵过程始终处于最优化状态,优化结果可提供给工艺工程师加以修正,再应用于生产,从而提高生产率。

Optimal software is designed for ferment process control and the advanced control theories are used in it. The neural network model is established to estimate biomass concentration, substrate concentration and produce concentration,so as to overcome the strong non-linearity in the ferment processes with time varying parameters. Although lacking the accurate mathematics model of the biotechnological processes, the neural network nonlinear prediction control method that absorbs the genetic algorithm can search for the optimal parameter in ferment process. The optimization results can be revised by the technologic engineers, and then be applied in the production. So the productivity is increased because the state of ferment process is optimal.